From: Matt Rice <ratmice@gmail.com>
To: pmuldoon@redhat.com
Cc: gdb-patches@sourceware.org
Subject: Re: [patch] [python] find_line_pc_range
Date: Tue, 05 Jul 2011 02:47:00 -0000 [thread overview]
Message-ID: <CACTLOFremXGUmj6Gk=EHtmtP5HYdVZM_f8f=oeJJrWQgho+6zA@mail.gmail.com> (raw)
In-Reply-To: <m3zkkus925.fsf@redhat.com>
[-- Attachment #1: Type: text/plain, Size: 3382 bytes --]
On Mon, Jul 4, 2011 at 2:13 AM, Phil Muldoon <pmuldoon@redhat.com> wrote:
> Matt Rice <ratmice@gmail.com> writes:
>
>> Thanks, attached is an updated patch that also includes tests.
>
> Thanks.
>
>> + ret_tuple = PyTuple_New (2);
>> + if (!ret_tuple)
>> + return Py_None;
>
> Return NULL here. This signifies an error, and that we have abandoned
> the function. When a Python function returns NULL, the exception is
> eventually printed and cleared either by the callers, or by the GDB
> Python top-level exception handling code.
Thanks for explaining, that is what I wanted. You said so in your first email,
but for some reason it didn't latch on, sorry.
>> + start = gdb_py_long_from_ulongest (start_pc);
>> + if (!start)
>> + goto fail_ret;
>> + if (PyTuple_SetItem (ret_tuple, 0, start) != 0)
>> + {
>> + Py_XDECREF (start);
>> + goto fail_ret;
>> + }
>
> You do not need to decrement "start" here. Even if the tuple SetItem
> call failed, you have already passed the responsibility to the tuple,
> and you no longer own it (stolen reference). SetItem would decrement
> this on failure. So just goto to your local error block.
OK, the docs were silent on this regard, and I guessed on error it
wouldn't steal it.
>> + end = gdb_py_long_from_ulongest (end_pc);
>> + if (!end)
>> + goto fail_ret;
>> + if (PyTuple_SetItem (ret_tuple, 1, end) != 0)
>> + {
>> + Py_XDECREF (end);
>> + goto fail_ret;
>> + }
>
> Same as above.
>
>> + fail_ret:
>> + Py_XDECREF (ret_tuple);
>
> Small nit, and this is generally ok, but you already know that if this
> error block is called, ret_tuple will be instantiated. So in this case
> the XDECREF is a tiny bit redundant. You could just use DECREF here and
> skip the NULL check that XDECREF does. Like I said, a tiny nit, but I
> just thought I would point it out.
k.
>> + return Py_None;
>
> If there is a Python exception that you cannot handle internally to your
> function, you must always return NULL. This informs callers, and
> ultimately the exception managing code in GDB that some Python call
> failed, to print the exception, and clear it.
>
>> @@ -526,6 +573,8 @@ static PyMethodDef sal_object_methods[] = {
>> { "is_valid", salpy_is_valid, METH_NOARGS,
>> "is_valid () -> Boolean.\n\
>> Return true if this symbol table and line is valid, false if not." },
>> + { "find_line_pc_range", salpy_find_line_pc_range, METH_NOARGS,
>> + "Return a tuple of start and end ranges for the symbol table." },
>> {NULL} /* Sentinel */
>> };
>
> We try to put the return type in the help now as a formal statement.
> Look at the two-line help example above: "is_valid()" for guidance.
Updated that text a little bit also, along with using Eli's text for the manual.
2011-07-04 Matt Rice <ratmice@gmail.com>
* python/py-symtab.c: Populate sal_object_methods.
(salpy_find_line_pc_range): New function.
2011-07-04 Matt Rice <ratmice@gmail.com>
Eli Zaretskii <eliz@gnu.org>
* gdb.texinfo (Symbol Tables In Python): Add find_line_pc_range method.
2011-07-04 Matt Rice <ratmice@gmail.com>
* gdb.python/py-symtab.exp: New Tests for find_line_pc_range.
[-- Attachment #2: foo.diff --]
[-- Type: application/octet-stream, Size: 3528 bytes --]
diff --git a/gdb/doc/gdb.texinfo b/gdb/doc/gdb.texinfo
index dbaf30e..e183693 100644
--- a/gdb/doc/gdb.texinfo
+++ b/gdb/doc/gdb.texinfo
@@ -23293,6 +23293,14 @@ exist in @value{GDBN} any longer. All other
@code{gdb.Symtab_and_line} methods will throw an exception if it is
invalid at the time the method is called.
@end defmethod
+
+@defmethod Symtab_and_line find_line_pc_range
+Attempt to find the range of program counter addresses for the
+@code{line} attribute of the @code{Symtab_and_line} object. If
+found, return a @code{Tuple} containing the start and end addresses
+for the line. Otherwise (e.g., a line with no corresponding code or
+not present in the debug info), return @code{None}.
+@end defmethod
@end table
A @code{gdb.Symtab} object has the following attributes:
diff --git a/gdb/python/py-symtab.c b/gdb/python/py-symtab.c
index 107cdec..5ae1975 100644
--- a/gdb/python/py-symtab.c
+++ b/gdb/python/py-symtab.c
@@ -215,6 +215,46 @@ salpy_get_line (PyObject *self, void *closure)
}
static PyObject *
+salpy_find_line_pc_range (PyObject *self, PyObject *args)
+{
+ struct symtab_and_line *sal = NULL;
+ CORE_ADDR start_pc, end_pc;
+
+ SALPY_REQUIRE_VALID (self, sal);
+
+ if (find_line_pc_range (*sal, &start_pc, &end_pc))
+ {
+ PyObject *ret_tuple;
+ PyObject *start;
+ PyObject *end;
+
+ ret_tuple = PyTuple_New (2);
+ if (!ret_tuple)
+ return NULL;
+
+ start = gdb_py_long_from_ulongest (start_pc);
+ if (!start)
+ goto fail_ret;
+ if (PyTuple_SetItem (ret_tuple, 0, start) != 0)
+ goto fail_ret;
+
+ end = gdb_py_long_from_ulongest (end_pc);
+ if (!end)
+ goto fail_ret;
+ if (PyTuple_SetItem (ret_tuple, 1, end) != 0)
+ goto fail_ret;
+
+ return ret_tuple;
+
+ fail_ret:
+ Py_DECREF (ret_tuple);
+ return NULL;
+ }
+
+ return Py_None;
+}
+
+static PyObject *
salpy_get_symtab (PyObject *self, void *closure)
{
struct symtab_and_line *sal;
@@ -526,6 +566,10 @@ static PyMethodDef sal_object_methods[] = {
{ "is_valid", salpy_is_valid, METH_NOARGS,
"is_valid () -> Boolean.\n\
Return true if this symbol table and line is valid, false if not." },
+ { "find_line_pc_range", salpy_find_line_pc_range, METH_NOARGS,
+ "find_line_pc_range () -> Tuple.\n\
+Return a tuple of start and end ranges for the symbol table's line,\n\
+or None if not found." },
{NULL} /* Sentinel */
};
diff --git a/gdb/testsuite/gdb.python/py-symtab.exp b/gdb/testsuite/gdb.python/py-symtab.exp
index c52f5ef..e27c066 100644
--- a/gdb/testsuite/gdb.python/py-symtab.exp
+++ b/gdb/testsuite/gdb.python/py-symtab.exp
@@ -58,6 +58,14 @@ gdb_test "python print sal.symtab" "gdb/testsuite/gdb.python/py-symbol.c.*" "Tes
gdb_test "python print sal.pc" "${decimal}" "Test sal.pc"
gdb_test "python print sal.line" "42" "Test sal.line"
gdb_test "python print sal.is_valid()" "True" "Test sal.is_valid"
+gdb_test "python print sal.pc == sal.find_line_pc_range()\[0\]" "True" "Test sal.find_line_pc_range() 1"
+gdb_test "python print sal.find_line_pc_range()\[1\]" "${decimal}" "Test sal.find_line_pc_range() 2"
+
+# Test sal find_line_pc_range errors.
+gdb_test "python bad_line_sal = gdb.decode_line(\"py-symbol.c:404\")\[1\]\[0\]" "" \
+"test decode_line py-symbol.c:404"
+gdb_test "python print bad_line_sal.find_line_pc_range() == None" "True" \
+"find_line_pc_range for non-existent line"
# Test symbol table.
gdb_test "python print symtab.filename" "testsuite/gdb.python/py-symbol.c.*" "Test symtab.filename"
next prev parent reply other threads:[~2011-07-05 2:26 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-07-02 4:03 Matt Rice
2011-07-03 16:35 ` Phil Muldoon
2011-07-04 5:38 ` Phil Muldoon
2011-07-04 6:36 ` Matt Rice
2011-07-04 10:09 ` Phil Muldoon
2011-07-05 2:47 ` Matt Rice [this message]
2011-07-05 20:41 ` Phil Muldoon
2011-07-04 10:50 ` Eli Zaretskii
2011-07-04 10:56 ` Matt Rice
2011-07-04 13:14 ` Eli Zaretskii
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to='CACTLOFremXGUmj6Gk=EHtmtP5HYdVZM_f8f=oeJJrWQgho+6zA@mail.gmail.com' \
--to=ratmice@gmail.com \
--cc=gdb-patches@sourceware.org \
--cc=pmuldoon@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox